On retrieving sea ice freeboard from ICESat laser altimeter

Abstract

Abstract. Sea ice freeboard derived from satellite altimetry is the basis for the estimation of sea ice thickness using the assumption of hydrostatic equilibrium. High accuracy of altimeter measurements and freeboard retrieval procedure are, therefore, required. As of today, two approaches for estimating the freeboard using laser altimeter measurements from Ice, Cloud, and land Elevation Satellite (ICESat), referred to as tie points (TP) and lowest-level elevation (LLE) methods, have been developed and applied in different studies. We reproduced these methods for the ICESat observation periods (2003)(2004)(2005)(2006)(2007)(2008) in order to assess and analyse the sources of differences found in the retrieved freeboard and corresponding thickness estimates of the Arctic sea ice as produced by the Jet Propulsion Laboratory (JPL) and Goddard Space Flight Center (GSFC). Three main factors are found to affect the freeboard differences when applying these methods: (a) the approach used for calculation of the local sea surface references in leads (TP or LLE methods), (b) the along-track averaging scales used for this calculation, and (c) the corrections for lead width relative to the ICESat footprint and for snow depth accumulated in refrozen leads. The LLE method with 100 km averaging scale, as used to produce the GSFC data set, and the LLE method with a shorter averaging scale of 25 km both give larger freeboard estimates comparing to those derived by applying the TP method with 25 km averaging scale as used for the JPL product. Two factors, (a) and (b), contribute to the freeboard differences in approximately equal proportions, and their combined effect is, on average, about 6-7 cm. The effect of using different methods varies spatially: the LLE method tends to give lower freeboards (by up to 15 cm) over the thick multiyear ice and higher freeboards (by up to 10 cm) over first-year ice and the thin part of multiyear ice; the higher freeboards dominate. We show that the freeboard underestimation over most of these thinner parts of sea ice can be reduced to less than 2 cm when using the improved TP method proposed in this paper. The corrections for snow depth in leads and lead width, (c), are applied only for the JPL product and increase the freeboard estimates by about 7 cm on average. Thus, different approaches to calculating sea surface references and different along-track averaging scales from one side and the freeboard corrections as applied when producing the JPL data set from the other side roughly compensate each other with respect to freeboard estimation. Therefore, one may conclude that the difference in the mean sea ice thickness between the JPL and GSFC data sets reported in previous studies should be attributed mostly to different parameters used in the freeboard-to-thickness conversion.

Quality assessment of the TOPAZ4 reanalysis in the Arctic over the period 1991–2013

“…Sea ice thickness TOPAZ4 shows some large biases (approximately −1.1 m) compared to ice thickness from ICESat and IceBridge as well as compared to ice draft data, although the thick ICESat ice draft may have been overestimated (Khvorostovsky and Rampal, 2016). The thickness bias is largest north of Ellesmere Island with bias up to 2 m. The spatial pattern and regression compare reasonably well.…”

“…Also the complementary thickness of thick ice from ICESat Khvorostovsky and Rampal, 2016) and CryoSat-2 (Wingham et al, 2006;Laxon et al, 2013), and SMOS sea surface salinity (Reul et al, 2012) will be tested in order to determine how to better assimilate into the system in the near future.…”

Impact of assimilating a merged sea-ice thickness from CryoSat-2 and SMOS in the Arctic reanalysis

“…A similar assessment over the same time frame has been carried out in the Arctic Cap Nowcast/Forecast System (ACNFS) by Allard et al (2018) revealing significant improvements of bias and root mean square difference (RMSD) but little changes in ice velocity except in marginal seas. The proposed study in complementary to Allard et al (2018) because the TOPAZ4 prediction system uses a more rudimentary sea-ice thermodynamics (no explicit ice thickness distribution) but a more advanced ensemble-based data assimilation method (TOPAZ4 uses strongly coupled data assimilation of ocean and sea ice, meaning that sea-ice observation will impact also the ocean and vice versa with a flow dependent assimilation method; see Penny et al, 2017;Kimmritz et al, 2018).…”

Augmenting in situ lake level measurements with Earth observation satellites

“…High accuracy (2~ 10 cm) of GLAS derived surface heights over flat regions are already mentioned in scientific literature [31][32][33]. Among various GLAS data sets, GLAS / ICESat L2 Global Land Surface Altimetry Data (GLA14) which provides surface heights over land, Version 33 is used.…”